How to recommend the best practice on Microsoft Azure


One of the directives of the One Commercial Partner (OCP) team at Microsoft is to help partners build net new practices and solution offerings using the Microsoft cloud. In our travels, we’re often asked, “What’s the best practice for …?” The answer is always, “It depends,” and frequently creates the opportunity for a long discussion about what it takes to create a solution architecture in the cloud. Since the topic is so relevant to so many of our partners, we wanted to distill it for you in this post.

What Microsoft Azure brings to the table

In years past, when we designed solutions for the ground we were always limited. Something constrained what we could do—financial, technical, and time-based constraints being the most common. With solution architecture in the cloud, many of those constraints have been eliminated, or at least greatly reduced. A fitting example of this is designing with high-availability (HA) in mind. On the ground, achieving HA typically meant implementing clustering. Due to the cost, complexity, and time, around 20% (or less) of customers actually implemented clustering.

With Microsoft Azure, the equivalent of clustering is Availability Sets, which as you know is simply a few fields to complete and you’re done. The complexity and time constraints are removed, and the cost constraint is greatly reduced. Microsoft Azure democratizes what used to be complex and costly technology solutions and makes them available to everyone, and that’s the first thing to keep in mind when designing for the cloud: Everything the cloud offers is available to you as you design your solution.

Best vs. recommended

What’s the best thing to do with all these options? Well, let’s first talk about the semantics of best vs. recommended.

These two words—and more specifically the subtle difference between them—can lead to trouble if they’re used interchangeably. In the consulting world, the best thing to do might not be the recommended thing to do, and vice versa, so let’s take a closer look at that by exploring the definition of each word.

Dictionary.com defines best as “of the highest quality, excellence, or standing: most advantageous, suitable, or desirable: largest; most:” Recommend is defined as “to present as worthy of confidence, acceptance, use, etc.; commend; mention favorably: to represent or urge as advisable or expedient: to advise, as an alternative; suggest (a choice, course of action, etc.) as appropriate, beneficial, or the like:” Mention favorably is the key differentiator. Best is desirable, whereas recommended is the thing to do.

We suggest that it’s actually better to stay away from using the phrase “best practices,” since what’s best for one customer or situation is not necessarily best for another customer or situation. Best is an adjective, recommend is a verb. Best is relative, recommend is specific. As consultants, you’re hired for the knowledge and experience you bring to the table (back to the verb thing again). Situations may change, and what was best yesterday may not be the best right now. Therefore, based upon XYZ requirements, you recommend this solution or approach. Let’s look at another example.

Everyone typically agrees that exercise is good for you; it’s a best practice. Well, what kind of exercise? Cardio? Weight training? HIIT? Yoga? If you visit a fitness coach, they’ll assess your situation and recommend what to do given all sorts of factors. The same concept carries over to the role of Cloud Solution Architects. Given a whole bunch of factors, you can make a recommendation on what to do.

With all this said, Microsoft is no exception in terms of needing to choose phrasing thoughtfully. All throughout our publications, we use the term “best practices.” Just remember that as you read or learn about a “best practice,” it’s important to understand what it is, what it does, and how you can leverage it. Most of all, it’s critical to remember that the details of its implementation may differ between situations. To succeed doing architecture in the cloud, you need to know what all the options are, seek to understand what your customer wants to do, and then make a recommendation.

What do you recommend?

In our experience working with hundreds of partners and customers, there isn’t one thing in Microsoft Azure that’s recommended to do in every situation with every customer. Some might say “security,” and that’s correct but broad. Rather, which specific security solutions are recommended vary from customer to customer.

Community call

Want to hear more? Join our Applications and Infrastructure Community call on Friday, June 15, 2018 at 10:00 a.m. PT. We’ll be talking about various compute, network, and storage capabilities in Microsoft Azure that are important to be familiar with as you design solutions in a conversational dialogue. Register here!

Applications and Infrastructure Technical Community

Skip to main content